includes/SearchPostgres.php

   1 <?php
   2 # Copyright (C) 2006-2007 Greg Sabino Mullane <greg@turnstep.com>
   3 # http://www.mediawiki.org/
   4 #
   5 # This program is free software; you can redistribute it and/or modify
   6 # it under the terms of the GNU General Public License as published by
   7 # the Free Software Foundation; either version 2 of the License, or
   8 # (at your option) any later version.
   9 #
  10 # This program is distributed in the hope that it will be useful,
  11 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  12 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13 # GNU General Public License for more details.
  14 #
  15 # You should have received a copy of the GNU General Public License along
  16 # with this program; if not, write to the Free Software Foundation, Inc.,
  17 # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18 # http://www.gnu.org/copyleft/gpl.html
  19
  20 /**
  21  * Search engine hook base class for Postgres
  22  * @addtogroup Search
  23  */
  24
  25 class SearchPostgres extends SearchEngine {
  26
  27         function SearchPostgres( $db ) {
  28                 $this->db = $db;
  29         }
  30
  31         /**
  32          * Perform a full text search query via tsearch2 and return a result set.
  33          * Currently searches a page's current title (page.page_title) and
  34          * latest revision article text (pagecontent.old_text)
  35          *
  36          * @param string $term - Raw search term
  37          * @return PostgresSearchResultSet
  38          * @access public
  39          */
  40         function searchTitle( $term ) {
  41                 $resultSet = $this->db->resultObject( $this->db->query( $this->searchQuery( $term , 'titlevector', 'page_title' )));
  42                 return new PostgresSearchResultSet( $resultSet, $this->searchTerms );
  43         }
  44         function searchText( $term ) {
  45                 $resultSet = $this->db->resultObject( $this->db->query( $this->searchQuery( $term, 'textvector', 'old_text' )));
  46                 return new PostgresSearchResultSet( $resultSet, $this->searchTerms );
  47         }
  48
  49
  50         /*
  51          * Transform the user's search string into a better form for tsearch2
  52         */
  53         function parseQuery( $term ) {
  54
  55                 wfDebug( "parseQuery received: $term" );
  56
  57                 ## No backslashes allowed
  58                 $term = preg_replace('/\\\/', '', $term);
  59
  60                 ## Collapse parens into nearby words:
  61                 $term = preg_replace('/\s*\(\s*/', ' (', $term);
  62                 $term = preg_replace('/\s*\)\s*/', ') ', $term);
  63
  64                 ## Treat colons as word separators:
  65                 $term = preg_replace('/:/', ' ', $term);
  66
  67                 $this->searchTerms = array();
  68                 $m = array();
  69                 $searchstring = '';
  70                 if( preg_match_all('/([-!]?)(\S+)\s*/', $term, $m, PREG_SET_ORDER ) ) {
  71                         foreach( $m as $terms ) {
  72                                 if (strlen($terms[1])) {
  73                                         $searchstring .= ' & !';
  74                                 }
  75                                 if (strtolower($terms[2]) === 'and') {
  76                                         $searchstring .= ' & ';
  77                                 }
  78                                 else if (strtolower($terms[2]) === 'or' or $terms[2] === '|') {
  79                                         $searchstring .= ' | ';
  80                                 }
  81                                 else if (strtolower($terms[2]) === 'not') {
  82                                         $searchstring .= ' & !';
  83                                 }
  84                                 else {
  85                                         $searchstring .= " & $terms[2]";
  86                                         $safeterm = preg_replace('/\W+/', '', $terms[2]);
  87                                         $this->searchTerms[$safeterm] = $safeterm;
  88                                 }
  89                         }
  90                 }
  91
  92                 ## Strip out leading junk
  93                 $searchstring = preg_replace('/^[\s\&\|]+/', '', $searchstring);
  94
  95                 ## Remove any doubled-up operators
  96                 $searchstring = preg_replace('/([\!\&\|]) +(?:[\&\|] +)+/', "$1 ", $searchstring);
  97
  98                 ## Remove any non-spaced operators (e.g. "Zounds!")
  99                 $searchstring = preg_replace('/([^ ])[\!\&\|]/', "$1", $searchstring);
 100
 101                 ## Remove any trailing whitespace or operators
 102                 $searchstring = preg_replace('/[\s\!\&\|]+$/', '', $searchstring);
 103
 104                 ## Remove unnecessary quotes around everything
 105                 $searchstring = preg_replace('/^[\'"](.*)[\'"]$/', "$1", $searchstring);
 106
 107                 ## Quote the whole thing
 108                 $searchstring = $this->db->addQuotes($searchstring);
 109
 110                 wfDebug( "parseQuery returned: $searchstring" );
 111
 112                 return $searchstring;
 113
 114         }
 115
 116         /**
 117          * Construct the full SQL query to do the search.
 118          * @param string $filteredTerm
 119          * @param string $fulltext
 120          * @private
 121          */
 122         function searchQuery( $term, $fulltext, $colname ) {
 123
 124                 $searchstring = $this->parseQuery( $term );
 125
 126                 ## We need a separate query here so gin does not complain about empty searches
 127                 $SQL = "SELECT to_tsquery('default',$searchstring)";
 128                 $res = $this->db->doQuery($SQL);
 129                 if (!$res) {
 130                         ## TODO: Better output (example to catch: one 'two)
 131                         die ("Sorry, that was not a valid search string. Please go back and try again");
 132                 }
 133                 $top = pg_fetch_result($res,0,0);
 134
 135                 if ($top === "") { ## e.g. if only stopwords are used XXX return something better
 136                         $query = "SELECT page_id, page_namespace, page_title, 0 AS score ".
 137                                 "FROM page p, revision r, pagecontent c WHERE p.page_latest = r.rev_id " .
 138                                 "AND r.rev_text_id = c.old_id AND 1=0";
 139                 }
 140                 else {
 141                         $query = "SELECT page_id, page_namespace, page_title, ".
 142                         "rank($fulltext, to_tsquery('default',$searchstring),5) AS score ".
 143                         "FROM page p, revision r, pagecontent c WHERE p.page_latest = r.rev_id " .
 144                         "AND r.rev_text_id = c.old_id AND $fulltext @@ to_tsquery('default',$searchstring)";
 145                 }
 146
 147                 ## Redirects
 148                 if (! $this->showRedirects)
 149                         $query .= ' AND page_is_redirect = 0'; ## IS FALSE
 150
 151                 ## Namespaces - defaults to 0
 152                 if ( count($this->namespaces) < 1)
 153                         $query .= ' AND page_namespace = 0';
 154                 else {
 155                         $namespaces = implode( ',', $this->namespaces );
 156                         $query .= " AND page_namespace IN ($namespaces)";
 157                 }
 158
 159                 $query .= " ORDER BY score DESC, page_id DESC";
 160
 161                 $query .= $this->db->limitResult( '', $this->limit, $this->offset );
 162
 163                 wfDebug( "searchQuery returned: $query" );
 164
 165                 return $query;
 166         }
 167
 168         ## Most of the work of these two functions are done automatically via triggers
 169
 170         function update( $pageid, $title, $text ) {
 171                 ## We don't want to index older revisions
 172                 $SQL = "UPDATE pagecontent SET textvector = NULL WHERE old_id = ".
 173                                 "(SELECT rev_text_id FROM revision WHERE rev_page = $pageid ".
 174                                 "ORDER BY rev_text_id DESC LIMIT 1 OFFSET 1)";
 175                 $this->db->doQuery($SQL);
 176                 return true;
 177         }
 178
 179         function updateTitle( $id, $title ) {
 180                 return true;
 181         }
 182
 183 } ## end of the SearchPostgres class
 184
 185
 186 class PostgresSearchResult extends SearchResult {
 187         function PostgresSearchResult( $row ) {
 188                 $this->mTitle = Title::makeTitle( $row->page_namespace, $row->page_title );
 189                 $this->score = $row->score;
 190         }
 191         function getScore() {
 192                 return $this->score;
 193         }
 194 }
 195
 196 class PostgresSearchResultSet extends SearchResultSet {
 197         function PostgresSearchResultSet( $resultSet, $terms ) {
 198                 $this->mResultSet = $resultSet;
 199                 $this->mTerms = $terms;
 200         }
 201
 202         function termMatches() {
 203                 return $this->mTerms;
 204         }
 205
 206         function numRows() {
 207                 return $this->mResultSet->numRows();
 208         }
 209
 210         function next() {
 211                 $row = $this->mResultSet->fetchObject();
 212                 if( $row === false ) {
 213                         return false;
 214                 } else {
 215                         return new PostgresSearchResult( $row );
 216                 }
 217         }
 218 }
 219
 220
 221 ?>